llama.cpp/ggml/include/ggml-cuda.h
slaren efdd713023
Some checks are pending
flake8 Lint / Lint (push) Waiting to run
more build fixes
2024-11-12 13:56:28 +01:00

48 lines
1.6 KiB
C

#pragma once
#include "ggml.h"
#include "ggml-backend.h"
#ifdef __cplusplus
extern "C" {
#endif
#ifdef GGML_USE_HIPBLAS
#define GGML_CUDA_NAME "ROCm"
#define GGML_CUBLAS_NAME "hipBLAS"
#elif defined(GGML_USE_MUSA)
#define GGML_CUDA_NAME "MUSA"
#define GGML_CUBLAS_NAME "muBLAS"
#else
#define GGML_CUDA_NAME "CUDA"
#define GGML_CUBLAS_NAME "cuBLAS"
#endif
#define GGML_CUDA_MAX_DEVICES 16
// backend API
GGML_BACKEND_API ggml_backend_t ggml_backend_cuda_init(int device);
GGML_BACKEND_API bool ggml_backend_is_cuda(ggml_backend_t backend);
// device buffer
GGML_BACKEND_API ggml_backend_buffer_type_t ggml_backend_cuda_buffer_type(int device);
// split tensor buffer that splits matrices by rows across multiple devices
GGML_BACKEND_API ggml_backend_buffer_type_t ggml_backend_cuda_split_buffer_type(int main_device, const float * tensor_split);
// pinned host buffer for use with the CPU backend for faster copies between CPU and GPU
GGML_BACKEND_API ggml_backend_buffer_type_t ggml_backend_cuda_host_buffer_type(void);
GGML_BACKEND_API int ggml_backend_cuda_get_device_count(void);
GGML_BACKEND_API void ggml_backend_cuda_get_device_description(int device, char * description, size_t description_size);
GGML_BACKEND_API void ggml_backend_cuda_get_device_memory(int device, size_t * free, size_t * total);
GGML_BACKEND_API bool ggml_backend_cuda_register_host_buffer(void * buffer, size_t size);
GGML_BACKEND_API void ggml_backend_cuda_unregister_host_buffer(void * buffer);
GGML_BACKEND_API ggml_backend_reg_t ggml_backend_cuda_reg(void);
#ifdef __cplusplus
}
#endif